SORT BY:

LIST ORDER
THREAD
AUTHOR
SUBJECT


SEARCH

IPS HOME


    [Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

    RE: iSCSI: aborting an immediate command with ABORT TASK




    The plug-in is meant to avoid having to resend the missing command.
    This is also the reason for mandating ABORT TASK on the same connection!

    Julo


    "Tony Battersby" <tonyb@cybernetics.com>
    Sent by: owner-ips@ece.cmu.edu

    04/09/02 16:11
    Please respond to tonyb

           
            To:        <Black_David@emc.com>
            cc:        <ips@ece.cmu.edu>
            Subject:        RE: iSCSI: aborting an immediate command with ABORT TASK

           


    > From: owner-ips@ece.cmu.edu [mailto:owner-ips@ece.cmu.edu]On Behalf Of
    > Black_David@emc.com
    > Sent: Tuesday, September 03, 2002 7:21 PM
    > To: tonyb@cybernetics.com
    > Cc: ips@ece.cmu.edu
    > Subject: RE: iSCSI: aborting an immediate command with ABORT TASK
    >
    >
    > Tony,
    >

    > > > 2) The race between an immediate command finishing normally
    > > > (causing the target to free the command and forget its
    > associated ITT)
    > > >  and the initiator aborting it

    >
    > That's not correct.  In case 2) the ABORT TASK always succeeds no
    > matter how the race turns out - SAM-2 Section 6.2 says:
    >
    >                  If the logical unit supports this function, a response of
    >                  FUNCTION COMPLETE shall indicate that the task was aborted
    >                  or was not in the task set.
    >
    > The words "or was not in the task set" are the crucial ones.  In
    > case 2), the ABORT TASK always returns FUNCTION COMPLETE, and
    > there may
    > or may not be a response from the aborted task prior to that FUNCTION
    > COMPLETE, but there MUST NOT be one afterwards.

    I am not really concerned about the response code to the ABORT TASK.  My
    concern is that the initiator may end up plugging a CmdSN hole when it
    shouldn't.  From Section 9.6.1 (draft 15),

    "...targets must consider the CmdSN received and return the "Function
    complete" response."

    I don't have a problem with returning the "Function complete" response.  I
    do have a problem with the target considering the CmdSN received if the
    initiator intended to abort an immediate command.

    Tony





Home

Last updated: Wed Sep 04 11:18:58 2002
11757 messages in chronological order